MMX idct_add for VP8.

Originally committed as revision 23886 to svn://svn.ffmpeg.org/ffmpeg/trunk

libavcodec/x86/vp8dsp-init.c

@@ -196,6 +196,7 @@ HVBILIN(ssse3,  8, 16, 16)
 extern void ff_vp8_idct_dc_add_mmx(uint8_t *dst, DCTELEM block[16], int stride);
 extern void ff_vp8_idct_dc_add_sse4(uint8_t *dst, DCTELEM block[16], int stride);
 extern void ff_vp8_luma_dc_wht_mmxext(DCTELEM block[4][4][16], DCTELEM dc[16]);
+extern void ff_vp8_idct_add_mmx(uint8_t *dst, DCTELEM block[16], int stride);
 #endif
 
 #define VP8_LUMA_MC_FUNC(IDX, SIZE, OPT) \
@@ -229,6 +230,7 @@ av_cold void ff_vp8dsp_init_x86(VP8DSPContext* c)
 #if HAVE_YASM
     if (mm_flags & FF_MM_MMX) {
         c->vp8_idct_dc_add                  = ff_vp8_idct_dc_add_mmx;
+        c->vp8_idct_add                     = ff_vp8_idct_add_mmx;
         c->put_vp8_epel_pixels_tab[0][0][0]     =
         c->put_vp8_bilinear_pixels_tab[0][0][0] = ff_put_vp8_pixels16_mmx;
         c->put_vp8_epel_pixels_tab[1][0][0]     =

libavcodec/x86/vp8dsp.asm

@@ -142,6 +142,9 @@ filter_h6_shuf1: db 0, 5, 1, 6, 2, 7, 3, 8, 4, 9, 5, 10, 6, 11,  7, 12
 filter_h6_shuf2: db 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6,  7, 7,  8,  8,  9
 filter_h6_shuf3: db 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8,  9, 9, 10, 10, 11
 
+pw_20091: times 4 dw 20091
+pw_17734: times 4 dw 17734
+
 cextern pw_3
 cextern pw_4
 cextern pw_64
@@ -923,6 +926,92 @@ cglobal vp8_idct_dc_add_sse4, 3, 3, 6
     pextrd  [r1+r2], xmm2, 3
     RET
 
+;-----------------------------------------------------------------------------
+; void vp8_idct_add_<opt>(uint8_t *dst, DCTELEM block[16], int stride);
+;-----------------------------------------------------------------------------
+
+; calculate %1=%2+%1; %2=%2-%1, with %3=temp register
+%macro SUMSUB 3
+    mova      %3, %1
+    paddw     %1, %2
+    psubw     %2, %3
+%endmacro
+
+; calculate %1=mul_35468(%1)-mul_20091(%2); %2=mul_20091(%1)+mul_35468(%2)
+;           this macro assumes that m6/m7 have words for 20091/17734 loaded
+%macro VP8_MULTIPLY_SUMSUB 4
+    mova      %3, %1
+    mova      %4, %2
+    pmulhw    %3, m6 ;20091(1)
+    pmulhw    %4, m6 ;20091(2)
+    paddw     %3, %1
+    paddw     %4, %2
+    psllw     %1, 1
+    psllw     %2, 1
+    pmulhw    %1, m7 ;35468(1)
+    pmulhw    %2, m7 ;35468(2)
+    psubw     %1, %4
+    paddw     %2, %3
+%endmacro
+
+; calculate x0=%1+%3; x1=%1-%3
+;           x2=mul_35468(%2)-mul_20091(%4); x3=mul_20091(%2)+mul_35468(%4)
+;           %1=x0+x3 (tmp0); %2=x1+x2 (tmp1); %3=x1-x2 (tmp2); %4=x0-x3 (tmp3)
+;           %5/%6 are temporary registers
+;           we assume m6/m7 have constant words 20091/17734 loaded in them
+%macro VP8_IDCT_TRANSFORM4x4_1D 6
+    SUMSUB_BA           m%3, m%1, m%5     ;t0, t1
+    VP8_MULTIPLY_SUMSUB m%2, m%4, m%5,m%6 ;t2, t3
+    SUMSUB_BA           m%4, m%3, m%5     ;tmp0, tmp3
+    SUMSUB_BA           m%2, m%1, m%5     ;tmp1, tmp2
+    SWAP                 %4,  %1
+    SWAP                 %4,  %3
+%endmacro
+
+; transpose a 4x4 table
+%macro TRANSPOSE4x4 5 ; output in %1/%4/%5/%3
+    mova      m%5, m%1
+    punpcklwd m%1, m%2
+    punpckhwd m%5, m%2
+    mova      m%2, m%3
+    punpcklwd m%3, m%4
+    punpckhwd m%2, m%4
+    mova      m%4, m%1
+    punpckldq m%1, m%3 ;col0
+    punpckhdq m%4, m%3 ;col1
+    mova      m%3, m%5
+    punpckldq m%5, m%2 ;col2
+    punpckhdq m%3, m%2 ;col3
+    SWAP       %4,  %2
+    SWAP       %4,  %5
+    SWAP       %4,  %3
+%endmacro
+
+INIT_MMX
+cglobal vp8_idct_add_mmx, 3, 3
+    ; load block data
+    movq         m0, [r1]
+    movq         m1, [r1+8]
+    movq         m2, [r1+16]
+    movq         m3, [r1+24]
+    movq         m6, [pw_20091]
+    movq         m7, [pw_17734]
+
+    ; actual IDCT
+    VP8_IDCT_TRANSFORM4x4_1D 0, 1, 2, 3, 4, 5
+    TRANSPOSE4x4W            0, 1, 2, 3, 4
+    paddw        m0, [pw_4]
+    VP8_IDCT_TRANSFORM4x4_1D 0, 1, 2, 3, 4, 5
+    TRANSPOSE4x4W            0, 1, 2, 3, 4
+
+    ; store
+    pxor         m4, m4
+    lea          r1, [r0+2*r2]
+    STORE_DIFFx2 m0, m1, m6, m7, m4, 3, r0, r2
+    STORE_DIFFx2 m2, m3, m6, m7, m4, 3, r1, r2
+
+    RET
+
 ;-----------------------------------------------------------------------------
 ; void vp8_luma_dc_wht_mmxext(DCTELEM block[4][4][16], DCTELEM dc[16])
 ;-----------------------------------------------------------------------------

libavcodec/x86/x86util.asm

@@ -365,3 +365,18 @@
     packuswb   %1, %1
     movh       %4, %1
 %endmacro
+
+%macro STORE_DIFFx2 8 ; add1, add2, reg1, reg2, zero, shift, source, stride
+    movh       %3, [%7]
+    movh       %4, [%7+%8]
+    punpcklbw  %3, %5
+    punpcklbw  %4, %5
+    psraw      %1, %6
+    psraw      %2, %6
+    paddw      %3, %1
+    paddw      %4, %2
+    packuswb   %3, %5
+    packuswb   %4, %5
+    movh     [%7], %3
+    movh  [%7+%8], %4
+%endmacro